Damon Linker’s recent column for The Week hits one of my favorite topics: how do you define religion? He starts off by taking a whack at what sounds like a silly piece in The New York Times:

One of Leo Strauss’ most illuminating essays begins with a provocation: “A social science that cannot speak of tyranny with the same confidence with which medicine speaks, for example, of cancer, cannot understand social phenomena as what they are. It is therefore not scientific.”

Something similar might be said about a religion columnist who finds it impossible to define religion.

I’m referring to The New York Times‘ Mark Oppenheimer, who penned a remarkable “Beliefs” column that ran on Saturday. The subject of the column? Whether CrossFit — a trendy form of demanding physical exercise — might be a form of religion. Oppenheimer’s answer? Sure! Because “it’s surprisingly hard to say what makes a religion.”

Is it really?

The fact that Oppenheimer found a woman who really, really likes CrossFit and thinks about it “as others might speak about a church or synagogue community” doesn’t prove that CrossFit really is her religion — any more than the fact that a couple of student researchers at Harvard Divinity School interviewed this same CrossFit fanatic for a study of “spaces other than churches that function as spiritual communities” demonstrates that a CrossFit gym is actually a place of religious worship.

Unless, of course, we define “religion” to mean “stuff someone really, really likes.”

Which is pretty much how Oppenheimer defines religion.

Already, I feel like saying “yes, but.” Clearly “stuff someone really, really likes” is inadequate as a definition of religion, but is also not really a fair description of this particular fanatic. If I said “CrossFit is her religion” you’d know what I meant – and that I didn’t just mean “she really likes CrossFit.” I’d mean: she goes to CrossFit without fail; her consciousness is perpetually occupied by CrossFit; in her personal Maslow’s hierarchy CrossFit is up at the top with her basic identity not down at the bottom with physiological needs.

Nonetheless, Linker is right that if I said “CrossFit is her religion” I wouldn’t have meant it literally, but as an analogy. I would have meant that she’s treating CrossFit as if it were a religion. So what is this thing “religion” that I’d be analogizing her behavior to?

Here’s Linker’s stab at a more serious definition:

Religion is any set of norms, practices, and beliefs that establishes a comprehensive way of life that is held out as the right or best way of life for those who adhere to it.

Many of these comprehensive ways of life posit the existence of one or more deities, but not all of them do — just as others teach that a life awaits us after death, while still others make no such claims. What matters is the comprehensiveness, not the content, of the way of life.

It is above all this comprehensiveness that precludes CrossFit from qualifying as a religion, even for those who take the fitness routine very seriously, because it is still just a form of physical exercise and not a sweeping statement of how a human being should live and understand his or her place in the universe. It makes no broader claims about the meaning or purpose of life, death, morality, love, and the origins, foundations, and ends of existence. The same holds for football, Star Trek, and dieting fads. Which is why those activities aren’t religions but Judaism, Christianity, Islam, Mormonism, Hinduism, Buddhism, Shintoism, Taoism, Jainism, and Sikhism very clearly are.

Again, I want to say “yes, but” even though I think Linker’s definition has something going for it – in particular, the idea of “comprehensiveness” seems crucial to what makes a religion, and what keeps CrossFit out of the realm of reasonable candidates for the title. Still, I would phrase things a bit differently. I would say that religion is a comprehensive set of normative practices that reflect or imply a set of beliefs about the nature of life and the right way to live it. Those beliefs may or may not be conscious, and may or may not be articulated and taught, in the way that the practices are.

Why do I want to make that amendment? Because I am fundamentally suspicious of definitions of religion that take Christianity as normative, and I think there’s a pronounced tendency to do precisely that. Religion is far older than Christianity; Christianity is far from typical of human religious practice and experience in many ways; and even Christianity is often misdescribed by privileging theology and dogma over the experience of adherents.

The word “religion” itself comes from a pre-Christian Roman word, and there is ancient dispute about what its origins are, whether it comes from a root meaning “to read over” (which would imply a connection to the ritual reading or reciting of texts, something still central to the practice of a vast array of religions) or from a root meaning “to bind back” (which would imply a connection to lifestyle prohibitions and restrictions that accompany a great many religions as well). It is the latter sense that predominated as the word entered English in the middle ages, where it originally referred to membership in a monastic order – the “religious” were contrasted not with the irreligious or atheistic but with the laity, those who did not live under a rule. Regardless, it’s clear that etymologically “religion” seems to have more to do with what one does than with what one believes, more to do with practice than with preaching.

Moreover, it’s a peculiarly Christian (or, more broadly, Abrahamic) conceit to center religion on revelation, when it is manifest that religion predates the age of the great revelatory religions. The ancient world clearly understood the category of religion, but their religions emerged organically from the mists of pre-history. And it would probably be wrong to say that they encompassed a set of beliefs about the right or best way to live life. Rather, they were the most comprehensive expression of what living life is, and of how we – Greeks, Romans, Egyptians, etc. – live it. That’s certainly how Herodotus would have understood it. Confucius, too, I suspect.

Some of the world’s existing religions became more systematized under pressure of contact with new religions that organized themselves more around a set of formal beliefs and authoritative teachings or revelations from a founder. Hinduism evolved this way after successive contact with Buddhism from within, Islam and Christianity from without. But that shouldn’t lead us to say that, before the advent of such religions, whatever humanity was up to wasn’t “really” or “fully” religious. Rather, it should make us question whether the features that distinguish religion for us – such as a founding revelation – aren’t really much less central to what makes something a religion qua religion than we think.

Linker goes on to contrast religion with philosophy, arguing that they both aim to be comprehensive but that they have different notions of truth. Philosophy, as Linker describes it, involves a quest for truth, whereas in religion truth is received (whether from revelation or tradition). I’m not convinced that the idea of talking about anything as a “rival” to religion is a particularly useful way of thinking. And I’m not sure that every philosophical system is quite as Socratic as Linker’s definition would imply.

So, for example: in the Western tradition, what are Stoicism or Epicureanism? They are philosophical systems that comprehensively address the right or best way of life, and I would question whether they are anchored in a relentless pursuit of the nature of truth. It seems to me these philosophies – unlike the endless Socratic quest – express answers more than they pursue questions. But they are philosophies, aren’t they? They aren’t religions, are they?

Or what about Pythagoreanism? The Pythagoreans plainly had views about religion – they had distinctive practices and ceremonies and so forth. But are they a distinctive religion? Or are the a particular philosophical approach to life that encompasses religion as it was understood within the Greek world? Kind of the way, say, Taoism is a philosophical approach to life that encompasses religion as it is understood within the Chinese world? But in that case, what’s left of the idea of philosophy as a rival to religion?

It seems to me that while Linker wants to make truth claims central to religion, and therefore to contrast religion with philosophy with regard to how they evaluate such claims, what’s really distinctive about religion is the centrality of the sacred as a category. It’s a category that is strikingly missing from Damon’s definition of religion – perhaps because it is not really accessible to philosophy.

So perhaps I would say that Linker’s definition of religion is one that suits the needs of philosophers who might understand themselves to be partisans of a rival approach to life.

That doesn’t mean it’s the best definition for understanding religion.